Prediction: NVIDIA Stock Has Returned 14,700% in 10 Years. Can It Do It Again?

NVIDIA reported fiscal Q2 FY27 revenue of $96.22 billion, up 105.85% year over year, and non-GAAP EPS of $2.22, marking its fifth consecutive earnings beat; Data Center revenue rose 117% to $89.02 billion. The company guided Q3 revenue to $108 billion with 74% gross margin, while management indicated supply can meet only about 70% of AI demand through fiscal 2028. The bullish $308.85 12-month target implies 36.55% upside from $226.18, but risks include no China data-center-compute revenue in guidance, $279 billion of supply commitments, $108.5 billion in cloud-partner guarantees, and gross margins expected to trough at 71%-72% in Q4.

Analysis

The relevant question is not whether NVDA can replicate its historical return, but whether the market is underpricing the shift from chip vendor to ecosystem financier. Longer receivables and partner guarantees effectively transfer part of the AI build-out's credit risk onto NVDA; this can sustain reported revenue through the next 1-3 quarters while raising the downside convexity if GPU-cloud utilization or funding conditions deteriorate. A miss in collections, rather than a demand shortfall, is the underappreciated trigger for a valuation reset because it challenges the quality of earnings and cash conversion simultaneously.

The supply ramp creates a second-order long in memory rather than an unqualified incremental NVDA long. High-bandwidth-memory and advanced-memory procurement commitments should tighten availability and preserve pricing for MU, while compressing gross margins for lower-priority system builders and enterprise-server vendors that lack allocation leverage. In the 6-18 month horizon, AVGO is the more credible competitive pressure on NVDA: hyperscalers pursuing custom silicon can reduce the addressable portion of accelerator spend even while total AI capex remains elevated, particularly if inference becomes a larger mix of workloads.

Consensus appears to treat excluded China revenue as pure upside optionality. The more material risk is that restricted product inventory and engineering resources are redirected into lower-margin compliant configurations or contested non-China markets, increasing competitive intensity versus AMD and custom ASICs. The bullish thesis is falsified by another material DSO increase, a guarantee/commitment step-up without matching operating cash flow, or hyperscaler capex guidance being cut; the next 1-3 months of supplier orders and cloud-partner funding are more important than headline GPU demand commentary.

Key Decisions for Investors

• Initiate a 3-6 month long NVDA / short AMD pair, sized beta-neutral: NVDA retains superior platform monetization while AMD must justify a substantially richer earnings multiple with sustained share gains. Target 15-20% relative outperformance; exit if AMD data-center growth accelerates relative to NVDA for two consecutive reports or NVDA guides material gross-margin deterioration beyond the transition period.
• Buy MU on 6-12 month weakness rather than chase NVDA after strong prints; the memory procurement cycle offers more direct upside to tight supply and a clearer operating-leverage setup. Risk is a rapid capacity response or accelerator deployment delays; reassess if HBM pricing or MU margin guidance fails to improve despite industry demand.
• Maintain an NVDA core only with a defined risk trigger: reduce exposure if DSO expands materially again or if operating cash flow fails to track earnings over the next two reports. Those metrics would indicate that partner financing, not end-demand, is carrying the revenue run rate.
• Use AVGO as the structural hedge against NVDA-specific platform risk over 6-18 months, preferably via a modest long AVGO overlay rather than a directional short NVDA. Custom-silicon design wins are slow to appear in quarterly numbers but can alter the terminal growth multiple before they materially affect revenue.